Traditionally, visual-based RGB-D SLAM systems only use correspondences with valid depth values for camera tracking, thus ignoring the regions without 3D ...
確定! 回上一頁